I hear ya. Though, using the 'win shares' formula as a measure of actual wins contributed (literally) invites subjectivity of it's own.

Regarding win shares, it certainly has the same limitations as all box score stats do (PER, SPM, etc.) But what it does is give you the most important aspects that go towards winning games, and over a large sample size (like the course of the season) those aspects dominate the "little things" that don't take place as often. Most of what is essential on offense is tracked by the box sore anyway - and even on defense, which obviously is hard to track in the box score (and also even by watching the games; assigning proper credit for defensive stops to players isn't clear-cut and can certainly differ depending on whom you talk to), finding defensive win shares over a large sample gives you a good idea of "who's who" defensively. It's not perfect (it misses on some guys like Joe Dumars), but then again that's not what the metric IS, anyway. That's why it's an estimate.

And once again, I'll take imperfect numbers over subjective criteria that can't be measured at all.

The key is capturing all the contributions in a game and attributing the to the player accordingly. You can't do ALL of it with box score stats; and you can with APM but you have to sift through alot of the statistical "noise" associated with your team's performance. So yes, people are still working on them. But they are valuable pieces of information to have, and they're great to use together.>>>>>>>>>>>>>>>>>>>>>>>>>>>

This is well stated. I guess I just am focusing on the limitations of the box score, which seems to be the staple resource for this endevour. Because of these limitations---I PERSONALLY am reticent to stamp 'win shares' as valid.... NOT because the information used to create the 'scores' isn't valuable----but because it ALONE doesn't really cannot truly determine 'wins contributed' (not literally). It's like a misnomer to me and instead of me seeing it as valuable, objective data----I frown at the attempt to then push the results to mean more than they do (actual number of team wins contributed to). I guess I just wish we weren't overreaching with what is otherwise VERY telling data.

I think I have seen the glossary here... but I recall there being some intimation that the 'WHOLE' formula was in the book you directed me to (Basketball on Paper) and NOT fully displayed in the glossary here (which disapppinted me).

What I'm looking for is the formula in black and white that can be used to transfer raw data from a box score to one's 'win share' figure.

The formula is actually found on this site under the Glossary. It is based on the information that is found in "Basketball on Paper".

"I'm not dismissing anything because of who I THINK should be there.

Why would you say such a thing?"

Not you per se, but alot of tend to do that in this sport - even if they're not aware of it.

"I just don't think (at the present) that you can validly estimate the # of wins contributed to a team's total by a player with a formula that plugs in personal statistics. To think you CAN...... might be irrational, frankly."

Not really. Obviously when you sit and watch your favorite team over the course of a season, you see that some players contribute more to their team's success than others. You can certainly show this empirically as well - players like LeBron, Wade, Dwight, Kobe, etc. are more important to their team winning games than players like Derek Fisher and Anderson Varejao. Certainly they put up their own numbers, but they're not producing in a vacuum - their numbers are directly correlated to what their teams do in a game. It's simply trying to figure out the average value that a point, offensive/defensive rebound, assist, etc. contributes to a win. In the case of adjusted +/-, it doesn't even use ANY box score stats; just point differential.

The key is capturing all the contributions in a game and attributing the to the player accordingly. You can't do ALL of it with box score stats; and you can with APM but you have to sift through alot of the statistical "noise" associated with your team's performance. So yes, people are still working on them. But they are valuable pieces of information to have, and they're great to use together.

And it certainly beats the childish, "ZOMG Kobe Bryant is the BEST because his scowl face just MAKES the Lakers win" line of reasoning.

Well, I don't want to dismiss them. I want to know more about them. I AM skeptical od advanced stats... but I need to examine them more in depth on an individual basis to be fair to each one. Is the 'win shares' formula in that 'Basketball on Paper' book you referred me to? I'm glad the one I found ISN'T 'the one.

I'm not dismissing anything because of who I THINK should be there.

Why would you say such a thing?

I just don't think (at the present) that you can validly estimate the # of wins contributed to a team's total by a player with a formula that plugs in personal statistics. To think you CAN...... might be irrational, frankly.

What does it it matter? Regardless what the author thinks of his own work, it is widely praised as an informative read in the APBRmetrics community.

"The title of the book, ironically, is one of the problems inherent with the belief (JMO) that wins con be derived simply by looking at a player's personal statistics in a vacuum.... 'Basketball on Paper'. I had to smile. All to often, 'things on paper' don't add up to what actually occurs."

Win shares (which by the way, doesn't use the formula you posted; it is alot more rigorous in its construction and employs reason) is simply an estimate; nothing more, nothing less. It uses empirical data and statistical methods to estimate how many wins a player contributes to his teams. And unlike people who are skeptical about the use of stats, WS and other metrics such as SPM aim for objectivity, not ridiculous and personal subjective standards of how ranks should be ranked. The authors behind the formulas don't claim they are perfect, nor do they tout them as the one "holy grail" formula that captures the game of basketball in a single number. But they are valid sources of evidence to use when evaluating player contribution.

Dismissing them because they don't fit your personal views of "who should be where" is entirely up to you, but to argue your own views based on subjective criteria that cannot be verified is irrational.

Now comparing Lebron's departure to Jordan's or making a comparission is not posible.

Jordan walked away from the best team in the league and current champions while Lebron got away from a team that clearly had no chance to even make the conference finals. This is why the departure of Lebron James will have a bigger negative impact on the Cavs compared to the Chicago Bull's 1994 season, which by the way, was a slightly improved well coached highly motived team with Pippen leading them. Of course this is not a negative thing on Jordan's legacy as the greatest ever even if he was only the fuel to push the Bulls 5 more wins into the postseason.

Jordan, Pippen, Bird, Magic, Malone, Barkley, etc. they all agree in their comments that back in their prime they thought differently regarding Lebron's desition.

Now lets look at the bigger picture, 5 years from now Lebron is aiming to win at least 3 rings, go back to the cavs and win there too!